'Market values strongly suggest a need for review. Especially now that competition from China is curtailed.' Bengt Holmström, MIT
'I fear Google is becoming the new old Microsoft, before the antitrust case. They may not be doing substantial harm now—but they could' David Autor, MIT
'The biggest economic challenge is the enormous scale economies that AI brings and the resulting further concentration of monopoly power. Will the[Big] Five be reduced to 4,3,2, or even 1?'[1] Jan Eeckhout, UCL
'Regulators should be vigilant and scrutinize practices going forward.' Pinelopi Goldberg, Yale
AI transforms the economy. It can now process complex inputs and generate human-like outputs, making it more versatile and scalable. Most top economists think AI like ChatGPT, has a measurable impact on national innovation,[2] and many agree that AI will substantially boost our per capita income over the next 30 years—perhaps more than the internet.[3],[4]
Globally, all firms now invest $175B per year in AI.[5] In 2023, AI firms will raise ~$50B ($14B of which for generative AI).[6]
Firms whose workers will use AI most—finance, information, management, etc.—had much higher stock returns due to the GPT launch.[7] Indeed, without these 'AI boom stocks,' US markets would be down 2%, not up 8% this year.[8] This corresponds to the market now expecting an extra ~$3.5T due to AI beyond what it already priced in—more than the UK's GDP (though part of this just shift to Big Tech e.g. an edutech firm that helps with homework lost users to ChatGPT and its share price dropped by >40%[9]). And despite aggressive US semiconductor export bans, Chinese tech stocks are also up ~30% this year, and all Chinese AI firms are building new AI models (though they are behind US AI firms).
So while gains from AI have spread across big firms so far, some argue Big Tech has no moat to build an AI-centered business strategy inside since open-source AI needs little compute and is ever more competitive.[10] While firms cannot protect their R&D perfectly and capture all the rents, because patents expire and industry know-how diffuses, long term, Big Tech will further concentrate the market. Big Tech firms have already profited most from AI R&D, which led to OpenAI's GPT. In 2023, MAMAA, Tesla and especially Nvidia stock (up by $200B to $1T—10x more than the value of OpenAI) drove 50% of the S&P 500 gains.[11]
Next we show how this trend will further intensify due to competition issues in M&A, economies of scope and scale, unfair advantages in terms of access to compute, data, and talent, and tax avoidance.